AIATSIS embraces AI to transform its information governance strategy

The challenge

A small team with limited records management experience and a legacy platform

AIATSIS has a small IT and records management team, who spend their time focusing on records digitization, so the team and the skillsets they bring were not necessarily built for ongoing records management. there was a legacy records management platform that was no longer fit-for-purpose.

AIATSIS needed a platform that would enable them to be compliant with its obligations as a government agency. This platform needed to ingest historical records, as well as new records created over time.

Limited data governance made ensuring compliance across the organization difficult, increasing compliance risk

Like many organizations in the initial stages of the COVID-19 pandemic, AIATSIS adopted new digital platforms to enable efficient remote work but struggled to make sure data and records created in these data sources were governed. There were difficulties with ensuring staff compliance, many data sources had unmanaged records.  

AIATSIS needed a centralized data inventory to ensure compliance across the data corpus, while shrinking the digital footprint by removing some legacy systems. This all needed to be achieved with minimal disruption to users' workflows, to preserve efficiency and productivity.

The RecordPoint solution

Federated data management centralizes governance

RecordPoint provided a single data trust platform offering centralized governance and search across relevant file repositories and business systems: Microsoft Teams and SharePoint Online.  

In-place management reduces overhead

Thanks to RecordPoint, the organization has been able to streamline its records management activities, freeing up users and enabling records and information managers to manage and dispose of records in-place. This reduces the compliance burden on the end-user and means they do not need to modify their workflows.

“We don't have to worry about whether they save things in the right spot, whether there's multiple documents floating around between people. If they all work in Microsoft Teams and in SharePoint we can catch it there,” ICT Operations Manager for AIATSIS Charles Mason says.

“When we were looking at the various options that we had, before we made the decision, it was really important that we weren't going to have to re-do everyone's workflow. But because we're saying to them ‘Look, if you just work here and save your work here, you don't have to do anything beyond that.’ It's an easy sell. We're not really making them do anything too onerous.”

Automated records classification improves compliance

RecordPoint provides automated records classification with a high-level of accuracy. The Classification Intelligence (CI) machine learning engine is built to help customers categorize and govern information at scale.

In combination with federated data management, CI allowed the organization to rapidly catalog and categorize its complete data corpus at a high level of accuracy.

The results

Successful adoption lowers overall compliance risk

While the team is about a third of the way through the rollout across the organization, Mason says there has already been a positive impact to the organization’s compliance posture.  

After some initial training on particular categories of records, Classification Intelligence is now running at between 95% and 100% accuracy, far better than what a human could manage. Part of this success came from how easy it was to secure buy-in from the organization.

“As someone who hasn't got a background in records and information management, it's a lot easier to understand and to explain to the rest of the organization what we're doing," Mason says.

“It was pretty straightforward, it’s easy for someone who knows about records, not about SharePoint.”

Automated records classification provides certainty and confidence data is captured and PII identified

Thanks to federated governance, in-place management, and Classification Intelligence, AIATSIS can have certainty that it has a comprehensive, accurately classified data inventory with any personally identifiable information (PII) identified, without a large amount of manual effort.

“The users are just now saving everything where they want to. And so not only is it really, really accurate, but it's also everything, there is nothing missing.” Mason says.

This contrasts with his experience of relying on end-users.

“History says, you might get really, really accurate users if they are well trained, and they are well engaged, but they will still miss things.”

Less manual effort for the records management team

Once the CI engine has been trained on a particular data source, the records management team can move to an audit role, periodically checking in to audit the results. This allows them to move on to higher value tasks, such as managing new data sources.  

“The ones that have been implemented are taking care of themselves now with a little bit of checking over. Grace Price, our Records Manager plays more of an audit role. She is now putting her concentrated work time into those new rollouts, whereas the ones that have been rolled out are taking care of themselves.”

‍